construction technologg laboratories= . . . ..

.

'' $EK;-;-. ~ ~ '.s. e.=

^' . . ,#4s; . . se-'' ; :. e ** _

+*% - - N,

. f .g-

Report to

FLUOR POWER SERVICES, INC. TENNESSEE VALLEY AUTHORITYChicago, Illinois Knoxville, Tennessee

DUKE POWER COMPANYCharlotte, N. C.

TESTS TO EVALUATECOEFFICIENT OF STATIC FRICTION

BETWEEN STEEL AND CONCRETE

by

B. G. Rabbat and H. G. Russell

.

Submitted byCONSTRUCTION TECHNOLOGY LABORATORIES

A Division of the Portland Cement Association5420 Old Orchard Road

Skokie, Illinois 60077February 1979

7902080A3&.

TESTS TO EVALUATE COEFFICIENT OF

STATIC FRICTION BETWEEN STEEL AND CONCRETE

by

B. G. Rabbat and H. G. Russell *

HIGHLIGHTS

An experimental investigation was conducted to determine

the coefficient of static friction betwean cast-in-place

concrete or grout and rolled steel plate. Push-off tests were

performed under conditions that represe t the interior and

exterior bearing surfaces of a containment vessel.

Test Program

Five sets of three similar push-off tests were conducted.

Specimens consisted of either concrete cast on top of, or grout

cast under a flat steel plate. Three sets of concrete speci-

mens were tested with a wet interface at normal stress levels

of 20, 60, and 100 psi. One set of concrete specimens was

tested with a dry interface at a normal stress of 60 psi. One

set of grout specimens was tested with wet interface at a nor-

mal stress of 60 psi. For each specimen, nominal bond stress

and coefficient of static friction at interface were measured.

Conclusions

Results of this investigation have shown the following:

*Respectively, Structural Engineer and Manager, StructuralDevelopment Section, Portland Cement Association, Skokie,Illinois

-1-

1. Bond strength for concrete specimens varied between 25

and 89 psi. For grout specimens, bond strength was

negligible.

2. For wet concrete interface specimens average effective *

coefficients of static friction were 0.67, 0.65, and

0.64, at normal stresses of 20, 60, and 100 psi,

respectively. The corresponding average peak *

coefficients of static friction were 0.70, 0.68 and

0.64, respectively.

3. For dry concrete interface specimens with a normal

stress of 60 psi, the average effective and average

peak coefficiends of static friction were 0.57 and

0.69, respectively.

4. For wet grout interface and normal stress of 60 psi,

both average effective and average peak coefficients

of static friction were 0.68.

INTRODUCTION

The purpose of this investigation was to determine exper-imentally the coefficient of static friction between cast-in-

place concrete or grout and rolled steel plate. For concrete

specimens, wet and dry interface conditions were tested. For

grout specimens, wet interface condition only was tested. The

work described in this report complie.e with Fluor Power Services

Inc. Purchase Order No. B-1394 dated January 31, 1978.

* Definitions of effective and peak coefficients of staticfriction are given under heading " Shear Stress - SlipRelationship".

-2-

EXPERIMEN'.aL PROGRAM

This section describes the test specimens, test variables,

manufacturing procedure, test setup, instrumentation, and

testing procedure.

Test Specimens

Test specimens represented conditio.1s at either steel-

concrete or steel-grout interface in the lower support region

of a containment vessel. To simplify testing, the friction

surface was simulated as a flat plare.

A test specimen is shown in Fig. 1. The test surface was

13.5 in, wide and 24 in, long. These dimensions give a test

area of 324 sq.in.

Test variables

Fifteen specimens were tested in tnis investigation. All

specimens were tested in sets of three. Test variables inclu6ud

1. Concrete blocks with wet and dry interface

2. Grout blocks with wet interface

3. Level of normal compressive stress

Experimental program as summarized in Table 1 comprised the

following three series.

Series I - Wet Concrete-to-Steel Interface

Series I consisted of nine tests. Three tests were per-

formed at each of three levels of normal compressive stress.

After breaking the bond, the concrete-to-steel interface was

intentionally saturated with water for the friction tests.

-3-

C~oncrete or Grout Block

/7 Steel PlateI/,/ /

' i 14 "

$| / /// /v

24" 7 |a/'

4 28" y,

.

Fig. 1 Test Specimen

-4-

TABLE 1 - EXPERIMENTAL PROGRAM

NormalTest Compressive Interface

Number Series Specimen * Stress Condition(psi)

1 I CWB-1 602 CWA-1 100 Concrete to Steel3 CWC-1 20 (we t)4 CWB-2 605 CWA-2 1006 CWC-2 207 CWB-3 608 CWA-3 1009 CWC-3 20

10 II CDB-1 60 Concrete To Steel11 CDB-2 60 (dry)12 CDB-3 60

13 III GWB-1 60 Grout to Steel14 GWB-2 60 (we t)15 GWB-3 60

* Specimen Identification is as follows:

First Letter: C is for Concrete, G is for GroutSecond letter: W is for Wet, D is for DryThird letter is the Load Level: A= 100 psi, B 60 psi,=

C = 20 psiLast digit is for first, second or third test of each set.

-5-

Series II - Dry Concrete-to-Steel Interface

Three tests were performed in Series II. Normal compres-

sive stress was held constant at an intermediate level. The

concrete-to-steel interface was not wetted after breaking the

bond.

Series III - Wet Grout-to-Steel Interface

Series III consisted of three tests at the same level of

compressive stress as in Series II. However, after breaking

the bond the grout-to-steel interface was completely wetted.

Manufacture of Specimens

Concrete and grout specimens were manufactured according to

different procedures discussed below. In all specimens, the

steel plate was as rolled SA 516 Grade 70 steel, one inch thick

and 14x28 in, in size.

Steel Plate Cleaning

The steel plate was cleaned in accordance with the Steel

Structures Council Specifications SSPC-SP2-63, Hand Tool

III *Cleaning Oil and grease were removed using acetone.

solvent. Rust scale was removed by hand hammering. Loose mill

scale was removed by hand wire brushing.

Concrete-to-Steel Specimens

After cleaning the steel plate, wood formwork was clamped

to the top of the steel plate. The concrete mix was designed

to yield a compressive strength between 3000 and 5000 psi at 28

* Number in parenthesis denote references listed at the end ofthe report.

-6-

days. The mix was made using Type I Portland Cement, Elgin

Sand and Elgin Gravel with a maximum aggregate size of 1-1/2 in.

Concrete was compacted using an internal vibrator. From

each batch, six 6x12 in. standard concrete cylinders weretaken. These cylinders were used to determine the concrete

compressive strength.

Concrete was cured for 7 days under plastic - re ting .

Formwork was then stripped and specimens kept i n t.. . . labora-

tory at a temperature of 73 F and a relative humidity of 50%until test time at 28 to 30 days.

Grout-to-Steel Specimens

Grout-to-steel specimens were prepared in a setup as shownin Fig. 2. The setup was designed to simulate field conditions

of placing grout at the exterior bearing surface of acontainment vessel. Formwork was clamped t7 the underside ofthe steel plate. The specimen was inclined during grouting.

The difference in height between the bottom horizontal edges ofthe grout block was 4 in.

The grout mix specifications called for

Cement: Type 2, ASTM C-150, no admixtures

Fine Aggregate: Natural Elgin Sand, ASTM C33

Cement to Sand katio: 1 to 2 by Volume

Water to Cement Ratio: 0.5 by weight

Flow of this mix was measured after five drops in three secondson a flow table.

Grout was placed through a funnel secured to the top of theright-hand side pipe shown in Fig. 2. This pipe had a 2 in.

-7-

- - - ,_

* ?] ' *[j,; 2 | 5 ' ' ~~4 9.-y, , ,

my-

. i , , . , 'c , Lj .-- JP.i 3-,. ",

?. $;.} pig %

'~ - i=;Y. .hf

,'

. D:. s:.N- * -

pjf''.$n zif~~f';I &

~

f-, ,,:'' ' ' ' '

-,

. .,y g . :. '- . :f'Nf5' .* $:

_g.r . - ''4[ ? d' :.m- je. |8p;8

%,A '.cf- j ,,

.

.s. ..m.., ,. . ;o ,t4

- ;.

|*

..,o F.sdej ', ' bt)- y, : ',

Y-) $ ,$''''' ,..

*

?)'

i-

-

- '

x 'y. o-

1 . ,9]f.-ap .' ,_

- <

' . ' s..R'~ -' 3: ,I N *.) -

' ~r - -;s. , , *

t ;.t ,%; __ -

wg

. ,

,

.. ~,. :. - _,

=- =a . , -_ _M :;- - - ,- ~ -

Setup for Preparation of Grout S ecimensP-

-8-

diameter. Because of its consistency, the grout filled the

form and flowed upward in the lef t-hand pipe. The height ofthe left-hand side pipe corresponded to 5-ft head measured from

the center of the grout block.

To permit bleeding of air trapped under the plate, holes

were provided around the top perimeter of the f orm. These

holes were sealed as soon as air bleeding was completed.

The tubing was sawed off af ter 7 days. The specimens were

then stripped and turned upside down. They were kept in the

laboratory at a temperature of 73 F and a relative humidity of50% until test time. Grout compressive strength was measured

using six 2-in. cubes.

Test Setup

The test setup is shown in the photograph of Fig. 3 andschematically in Fig. 4. This setup is capable of applying

vertical loads normal to the shear plane and horizontal loads

parallel to the shear plane. The main components of the setup

are shown in Fig. 5. They were assembled as follows:

Test specimen was grouted to the lower block shown in Fig.5. Brackets clamped to the lower block served as guides toposition the steel plate. During testing, the brackets pre-

vented slip of the test specimen with respect to the lower

bloc k .

The upper block shown in Fig. 5 was grouted on top of thetest spec imen. Brackets embedded in the upper block served as

guides to place it. During testing, the brackets also helped

prevent slip of the specimen with respect to the upper block.

-9-

75 w aw fg

'

.

44

:

.

-\ n,.t m

*

a.

4 'l,

A 4

C'

,,

s *, ,' ' -

5 * ' le,_*

'M 1potentio ig!

. >* *? 5;

. ,.

, >, ,n

. .* -

*m yy ..,,

%,

, ~* ~e, %,:

sem ,s

...~-

.

l

S,M .' .-

.- - . *N . -

-10-

,-*-C B* A*,

A a E dp Load Cells n i-H- it-

1 ['

* Rams for Normal Stress

f h p -Structural Tubes ||'

Potentiometers

4- Load Cells

e /,

'I- i L__ ]_

||# 6|,|%, 'N x */ 11 ||

Rams for 11 - 11.Push-Off 'Y ' Y Y*Load LO B* AHALF-SECTION A-A HALF-SECTION B-B SECTION C-C

Fig. 4 Schematic of Test Setup

-*-B A+

UPPER BLOCK- _t ,, --

I"|+-- 24 /g H

"24" i| |*- 13 '/ M2

MConcrete Block TEST SPECIMEN ',',

SA 516 Grade 70) T,,Steel Plate 2 8,, _*

?.8 '/ " ~8

LOWER BLOCK

-+ B A+SECTION AA SECTION BB

Fig. 5 Test Setup Components

Horizontal rams were used to apply in-plane shear loads.

Vertical rams were used to apply normal stresses. All ramswere loaded using hydraulic oil pressure.

Instrumentation

All loads were measured using calibrated load cells }

acting in series with the hydraulic rams.

Slip of the concrete test block relative to the steel platewas measured with two potentiometers as shown in Fig. 6. Eachpotentiometer was attached to a bracket secured to the steel

plate. The potentiometer's plunger was attached to a bracket

on the concrete test block.

An X-Y plotter was used to obtain a continuous record of

the in-plane shear versus slip at the friction interface. The

Y-axis of the plotter was calibrated to indicate the in-plane, shear stress. This consisted of the sum of loads from two loadcells divided by the interface area. The X-axis indicated theaverage slip measured by two potentiometers.

Test Procedure

Bond between concrete and steel was broken by application

of an in-plane shear force applied at the shear-friction inter-face. No normal compressive stress other than the weight of

the upper block was applied at this stage. This weight wasequivalent to a nominal stress of 8.18 psi. The shear stressrequired to cause first slip was noted. The concrete block was

then moved back to its original position on the steel plate.

-13-

Concrete orGrout Block

._ Brocket securedto Block

p Potentiometer.

IJ /\ Steel Brocket secured to

Plate Steel Plate

Fig. 6 Attachments to Measure Slip

-14-

For tests requiring a wet friction surface, the iaterfaces

were separated af ter bond was broken. The interfaces were then

saturated with water to ensure a wet sur f ace during testing.

Interfaces remained saturated during subsequent testing.

During each test, normal stress was held constant.

Af ter application of normal compressive f orce, the shear

load was applied slowly. A continuous record of the horizontal

load versus slip at the shear-friction interface wac obtained.

A representative shear stress versus slip curve is shown in

Fig. 7. It can bee seen that slip occurred in increments.

Each slip increment occurred within a fraction of a second and

was accompanied by a drop in shear stress. The drop was a

result of the stif f ness of the loading system. To eliminate

the eff ect of stiff ness of the loading system, the shear stress

versus slip curves reported under test results were idealized

by a curve joining the peaks of the saw teeth of Fig. 7. Note

tha t the testing rate was adjusted to obtain 0.1 in. slip

within three to five minu tes. The test was stopped when 0.5

in. slip was reached.

TEST RESULTS

Results presented in this section include material

properties, bond strengchs at interface, shear stress versus

slip relationships, and coefficients of static f riction. The

eff ects of normal stresses, wet interface, grout versus

concrete, and concrete strength on the coefficient of static

f riction are discussed.

-15-

.

60 -

. #1NNS\\\iN\\\\\\\\\\\'4 ~ \Shear N \ .Stress,psi

20 -

O ' '

O 0.1 0.2

Slip, inch

Fig. 7 Representative Continuous Record of ShearStress versus Slip

-16-

Concrete Strength

At test age, six concrete cylinders were loaded to deter-

mine concrete compressive strength. The average concrete

strength is listed in Table 2. Age of concrete at time of

testing is also listed.

Grout Properties

Grout strength was measured using 2 in. cubes. The average

cube strength is listed in Table 3. The flow measured on the

flow table after five drops in three seconds is also listed.

Bond Strength at Interface

Bond strength is the nominal shear stress required to break

the bond between the concrete or grout block and the steel

plate. It is equal to the in-plane shear force that broke the

bond divided by the area of the interface (324 sq. in.).

Measured bond strengths are listed in Table 4 where the

specimens have been grouped in sets of three similar tests.

For concrete specimens, the bond strength varied between 25 and

89 psi.

For grout specimens, bond strength was negligible.

Inspection of the interface after testing indicated that after

placing the grout, bleeding occurred inside the formwork. Sand

particles settled and water laitance moved up, with a larger

accumulation at the highest point of the formwork. As cement

hydrated, water was absorbed. As a result, air gaps formed

between the steel plate and grout block. Moreover, air bubbles

-17-

TABLE 2 - CONCRETE STRENGTH

AverageTest Test Age Concrete Strength

Number Specimen (days) (psi)

1 CWB-1 28 37702 CWA-1 29 35103 CWC-1 28 32754 CWL-2 28 35105 CWA-2 28 37806 CWC-2 28 35257 CNB-3 28 30308 CWA-3 28 38009 CWC-3 28 3270

10 CDB-1 28 362011 CDB-2 28 372012 CDB-3 28 3710

TABLE 3 - GROUT PROPERTIES

Test Specimen Test Age Average CubeNumber (days) Days Strength (psi) Flow *

13 GWB-1 28 7680 14214 GWB-2 30 7655 12115 GWB-3 28 8390 120

1

* ASTM C230 (3)

.

-18-

TABLE 4 - BOND STRENGTH AND COEFFICIENT OF FRICTION

Bond Shear Stress (psi) Coefficient of FrictionSpecimen Strength

(psi) Peak Effective Peak Effective

CWA-1 52.8 62.0 62.0 0.62l 0.62CWA-2 48.6 63.5 63.5 0.63 0.64* 0.63 0.64CWA-3 51.4 68.0 67.8 0.681 0.68

CWB-1 55.0 42.0 40.7 0.70) 0.68CW2-2 25.0 40.4 39.6 0.67 ( 0.68 0.66 0.65CWB-3 89.0 40.5 36.1 0.68? 0.60

CWC-1 61.2 13.8 12.3 0.69| 0.64iCWC-2 83.6 14.2 14.0 0.71| 0.70 0.70f 0.67CWC-3 78.4 13.9 13.5 0.69? 0.671

CDB-1 67.3 40.5 33.7 0.68 0.56lCDB-2 53.0 46.2 35.0 0.77 0.69 0.58 0.57-CDB-3 58.2 36.9 34.0 0.62 0.57l

GWB-1 8.0 41.2 41.2 0.69l 0.69

0.68|1GWB-2 - 40.9 40.9 0.68 0.68 0.68GWB-3 - 40.0 40.0 0.67 0.67

* Average of each set

.

-19-

.

had formed at the interface as shown in Fig. 8. The left hand

side was the highest point as cast. In Specimens GWB-2 and

GWB-3, the bond had broken during handling of the specimens

before testing.

Shear Stress - Slip Relationship

Shear stress versus slip curves for Series I, II, and III

are plotted in Figs. 9, 10, and 11, respectively.

Peak and effective shear stresses are listed in Table 4.Peak shear stress is defined as the highest shear recorded on

shear stress versus slip curve. Where the peak shear stress

occurred at a slip less than 0.5 in., an effective shear stress

was determined. Effective shear stress is defined as the

lowest shear stress occurring after the peak shear stress.

Where the peak shear stress occurred at 0.5-in, slip, the

effective shear stress was considered equal to the peak shear

stress.

Coefficient of Friction

Coefficient of frictinn is defined as the shear stress

divided by the normal stress. Coefficients of friction

corresponding to each of the peak and effective shear stresses

are listea in Table 4. An average coefficient of friction for

each set is given in the same table.

-20-

f|f ',, kj, , .j' /;; - | i

'

,.. ~, ,1 j !!; : .i.

.,u

(' L*5 + =7w-

1

1

[' ,,!

_. ,_

...

._

.,

. _ ..

~- -- ,__

'N '. - N 3j ,, n D .'' s bl*.f.ld?i __. 2 1

Fig. 8 Friction interface ofSpecimen GiB-1 af ter Testing

-21-

i

80 -

NormalCompressive Stress

, _ _ . . . CWA.3100 psi

.- - -.. .-

.

A-2f CWK-l

~ ~

60 _ _ _ _ -/ p , _ ___ _ShearStress,psi

CWB-l 60 psi ,^

=7[N.~40 _ --CWB-2- '.s . _ _ _ _ . . _ . _ . , '0, CWB-3

20 -

--

FCWC-2 CWC-3 2 20 psi'

r ___.- -

L CWCT ~~-~

I iO0.1 0.2 03 o4 0.5

Slip, inch.

Fig. 9 Shear-Stress versus Slip of Series I

80 -

60 -

ShearStress,

.

-

40 J__

s ive Stressi / 60 ps1

EDh3''

' " * - ~ ~- - - - -

=a--

,

20 -

'' ' ' '0O 0.1 0.2 0.3 0.4 0.5

Slip, inch.

Fig. 10 She ar-S tre s s versus Slip of Series II

80-

60 -

SlieorStress, Normal

" "'" ""PSI 00 psiM B-3 GWB-l40

f_Y2- -

t

20 -

Oo'4

,

l O.2 03 0.5Si,lp, inch.

,

Fig ,11 Shear-Stress versus Slip Of Series III

Effect of Normal Stresses

The effect of normal stress on coefficient of friction is

shown in the results of Series I, concrete-to-concrete

specimens with wet interface. At normal stresses of 20, 60,

and 100 psi, the average peak coefficients of static friction

were 0.70, 0.68 and 0.64, respectively. The corresponding

average effective shear stresses were 0.67, 0.65, and 0.64,

respectively. These average coefficients of static friction

tended to decrease as normal compressive stress increased.

Effect of Wetting Interface

The effect of wet interface on coefficient of static

friction can be seen when comparing average results of CWB and

CDB tests in Table 4. Corresponding to peak shear stress, the

average coefficients of static friction are 0.68 and 0.69 for

wet and dry interfaces, respectively. These coefficients are

fairly close. However, the average coefficients corresponding

to effective shear stress, are 0.65 and 0.57 for wet and dryinterfaces, respectively. Therefore, it appears that the

'verage coefficient of friction corresponding to effective

.thear stress is about 12% lower for the case of dry interface.

Grout Versus Concrete Interface

Comparisons between coef ficient of static friction for

grout or concrete against rolled steel is obtained from the GWB

and CWB sets of specimens. The average peak coefficient of

static friction was the same in both cases a1though in the

-25-

grout series not all of the interface area was in contact with

the steel p_ ate. This interface condition with air gaps and

bubbles can be expected in the field. However, variation of

the grout surface condition did not appear to affect the

measured coefficient of friction.

The average effective coefficient of static friction for

concrete and grout specimens were 0.65 and 0.68, respectively.

The observed coefficient for the grout specimens was about 5%

higher than for the concrete specimens.

Effect of Concrete Strength

Concrete strengths are listed in Table 2, and coefficients

of static friction are listed in Table 4. Based on the test

results of Series I, the coefficient of static friction appar-

ently increased with increased concrete strength. However, the

range of concrete strengths is tco small to state positivelywhat the effect of concrete strength is on the coefficient of

static friction.

Concluding Remarks

Fif teen specimens were tested to determine the coefficient

of static friction between cast-in-place concrete or groutedmortar and rolled steel plate. The specimens simulated the

condition that could exist at the interior and exterior bearingsurfaces of a containment vessel. A detailed summary of the

test program, and conclusions appear under " HIGHLIGHTS" at the

beginning of this report.

-26-

ACKNOWLEDGMENTS

This investigation was carried out in the Structural

Development Section of the Construction Technology Laboratories

under Dr. W. G. Corley, Director, Engineering Development

Department. Particular credit is due to W. T. Fasig and W.

Hummerich, Jr. for their assistance in manufacture and testing

of the specimens.

CAVEAT

This report has been prepared for the exclusive use of

Fluor Power Services, Inc., Duke Power Company, and Tennessee

Valley Authority and their related entities. The use of the

report by others will be permitted only on the understanding

that there are no representations or warranties, expressed or

implied as to the accuracy and validity of the data and

information or conclusions contained therein.

-27-

REFERENCES

1. " Surface Preparation Specifications - No.2 Hand ToolCleaning", SSPC-SP2-63, Steel Structures Painting Council,October 1963.

2. Hognestad, E., Hanson, N.W., Kriz, L.B. and Kurvits, O.A.," Facilities and Test Methods of PCA Structural Laboratory",Journal of the PCA Research and Development Laboratories,Vol. 1, No. 1, pp. 12-20 and pp. 40-45, January 1959; Vol.1, No . 2, pp. 30-37, May 1959; Vol. 1, No. 3, pp. 35-41,No. 3, pp. 35-41, September 1959. Reprinted as DevelopmentDepartment Bulletin D33, Portland Cement Association,Skokie, Illinois.

3. Standard Specification for Flow Table for Use in Tests ofHydraulic Cement", ASTM C230-74, American Society forTesting and Materials, Philadelphia.

-28-